I only follow a small number of tags, but the articles in my feed seem to be only modestly influenced by these settings.

Actually, I don’t care as much about these particular tags as I do about the distinction between technical and nontechnical.

If someone writes a technical article about something I've never heard of, there's a decent chance I'd be interested in it, so I don't really want the tags I follow to be too powerful.

If I were doing it, I'd consider some kind of filter, probably based on machine learning. I think that the difference between technical and non-technical articles is sufficiently clear that an algorithm should be quite capable of making the determination. It doesn't have to be perfect, just good enough to shift the balance in the way the user wants.

code of conduct - report abuse